Elevation-Aware Domain Adaptation for Sematic Segmentation of Aerial Images

Recent advancements in Earth observation technologies have accelerated remote sensing (RS) data acquisition, yet cross-domain semantic segmentation remains challenged by domain shifts. Traditional unsupervised domain adaptation (UDA) methods often rely on computationally intensive and unstable gener...

Full description

Saved in:
Bibliographic Details
Main Authors: Zihao Sun, Peng Guo, Zehui Li, Xiuwan Chen, Xinbo Liu
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/17/14/2529
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849418056347942912
author Zihao Sun
Peng Guo
Zehui Li
Xiuwan Chen
Xinbo Liu
author_facet Zihao Sun
Peng Guo
Zehui Li
Xiuwan Chen
Xinbo Liu
author_sort Zihao Sun
collection DOAJ
description Recent advancements in Earth observation technologies have accelerated remote sensing (RS) data acquisition, yet cross-domain semantic segmentation remains challenged by domain shifts. Traditional unsupervised domain adaptation (UDA) methods often rely on computationally intensive and unstable generative adversarial networks (GANs). This study introduces elevation-aware domain adaptation (EADA), a multi-task framework that integrates elevation estimation (via digital surface models) with semantic segmentation to address distribution discrepancies. EADA employs a shared encoder and task-specific decoders, enhanced by a spatial attention-based feature fusion module. Experiments on Potsdam and Vaihingen datasets under cross-domain settings (e.g., Potsdam IRRG → Vaihingen IRRG) show that EADA achieves state-of-the-art performance, with a mean IoU of 54.62% and an F1-score of 65.47%, outperforming single-stage baselines. Elevation awareness significantly improves the segmentation of height-sensitive classes, such as buildings, while maintaining computational efficiency. Compared to multi-stage approaches, EADA’s end-to-end design reduces training complexity without sacrificing accuracy. These results demonstrate that incorporating elevation data effectively mitigates domain shifts in RS imagery. However, lower accuracy for elevation-insensitive classes suggests the need for further refinement to enhance overall generalizability.
format Article
id doaj-art-4edc190922c046a6ac5c070eb976c91e
institution Kabale University
issn 2072-4292
language English
publishDate 2025-07-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-4edc190922c046a6ac5c070eb976c91e2025-08-20T03:32:33ZengMDPI AGRemote Sensing2072-42922025-07-011714252910.3390/rs17142529Elevation-Aware Domain Adaptation for Sematic Segmentation of Aerial ImagesZihao Sun0Peng Guo1Zehui Li2Xiuwan Chen3Xinbo Liu4Institute of Remote Sensing and Geographic Information System, Peking University, Beijing 100871, ChinaLand Satellite Remote Sensing Application Center, Ministry of Natural Resources of P.R. China, Beijing 100048, ChinaInstitute of Remote Sensing and Geographic Information System, Peking University, Beijing 100871, ChinaInstitute of Remote Sensing and Geographic Information System, Peking University, Beijing 100871, ChinaChina Yangtze Power Co., Ltd., Yichang 443000, ChinaRecent advancements in Earth observation technologies have accelerated remote sensing (RS) data acquisition, yet cross-domain semantic segmentation remains challenged by domain shifts. Traditional unsupervised domain adaptation (UDA) methods often rely on computationally intensive and unstable generative adversarial networks (GANs). This study introduces elevation-aware domain adaptation (EADA), a multi-task framework that integrates elevation estimation (via digital surface models) with semantic segmentation to address distribution discrepancies. EADA employs a shared encoder and task-specific decoders, enhanced by a spatial attention-based feature fusion module. Experiments on Potsdam and Vaihingen datasets under cross-domain settings (e.g., Potsdam IRRG → Vaihingen IRRG) show that EADA achieves state-of-the-art performance, with a mean IoU of 54.62% and an F1-score of 65.47%, outperforming single-stage baselines. Elevation awareness significantly improves the segmentation of height-sensitive classes, such as buildings, while maintaining computational efficiency. Compared to multi-stage approaches, EADA’s end-to-end design reduces training complexity without sacrificing accuracy. These results demonstrate that incorporating elevation data effectively mitigates domain shifts in RS imagery. However, lower accuracy for elevation-insensitive classes suggests the need for further refinement to enhance overall generalizability.https://www.mdpi.com/2072-4292/17/14/2529unsupervised domain adaptationsemantic segmentationremote sensing imageself-supervisionmulti-task learning
spellingShingle Zihao Sun
Peng Guo
Zehui Li
Xiuwan Chen
Xinbo Liu
Elevation-Aware Domain Adaptation for Sematic Segmentation of Aerial Images
Remote Sensing
unsupervised domain adaptation
semantic segmentation
remote sensing image
self-supervision
multi-task learning
title Elevation-Aware Domain Adaptation for Sematic Segmentation of Aerial Images
title_full Elevation-Aware Domain Adaptation for Sematic Segmentation of Aerial Images
title_fullStr Elevation-Aware Domain Adaptation for Sematic Segmentation of Aerial Images
title_full_unstemmed Elevation-Aware Domain Adaptation for Sematic Segmentation of Aerial Images
title_short Elevation-Aware Domain Adaptation for Sematic Segmentation of Aerial Images
title_sort elevation aware domain adaptation for sematic segmentation of aerial images
topic unsupervised domain adaptation
semantic segmentation
remote sensing image
self-supervision
multi-task learning
url https://www.mdpi.com/2072-4292/17/14/2529
work_keys_str_mv AT zihaosun elevationawaredomainadaptationforsematicsegmentationofaerialimages
AT pengguo elevationawaredomainadaptationforsematicsegmentationofaerialimages
AT zehuili elevationawaredomainadaptationforsematicsegmentationofaerialimages
AT xiuwanchen elevationawaredomainadaptationforsematicsegmentationofaerialimages
AT xinboliu elevationawaredomainadaptationforsematicsegmentationofaerialimages